The ultimate goal of this proposal is to understand the genetic basis of the metabolic syndrome, a cluster of phenotypes that includes type 2 diabetes (T2D), obesity, hypertension, and dyslipidemia. These phenotypes account for a disproportionate amount of the public health burden in the US. Evolutionary genetics offers a powerful framework for investigating the metabolic syndrome because the risk genotypes are hypothesized to be result of metabolic adaptations to the diverse environments of ancestral human populations. In this multi-investigator proposal, we will address population genetics hypotheses in the following specific aims: 1. We will develop statistical methodology to test correlations between SNP allele frequencies and environmental variables against a null model of the correlations between populations due to population history. We will also select approx. 200 genes involved in the biological processes underlying the metabolic syndrome and 300 unconstrained genomic regions to be used as 'controls'. A total of 3072 SNPs - divided into approx. 2700 tag SNPs in the candidate genes and 300 SNPs in control regions (1 SNP/control region) -will be genotyped in 1056 individuals from 52 worldwide populations. The newly developed statistical methodology will be used to analyze the SNP genotyping data. 2. A set of approx. 20 candidate genes will be re-sequenced in 16 individuals each from 3 populations of African, European and Asian origin, respectively. The re-sequencing data will be analyzed to determine if a signature of positive natural selection is present in these genes and estimate the age of the selected alleles. 3. We will develop a formal model for the evolution of the metabolic syndrome in which disease risk is due to ancestral alleles that were maintained by purifying selection in ancient human populations and became either neutral or deleterious after the switch to the Western lifestyle. Population genetics simulations of this model will be performed to characterize the expected patterns of disease variation and linked neutral variation. The results of our study are likely to identify novel genetic variants that may affect the risk to the metabolic syndrome and will, more generally, help in the efficient design of disease association studies and in the interpretation of the results of studies aimed at replicating the original associations.